Compound-Signal Detection against an Unknown-Intensity Interference under Parameteric and Nonparameteric Prior Uncertainty

摘要

The detection of a signal which is the superposition of a deterministic and random component is studied in terms of the ε -contaminated model of the input random process. Decision rules (DRs) are synthesized using the least favorable distribution for the model. Scale invariance and the adaptive Bayesian approach are also employed.